In a recent interview, Google Deepmind’s CEO Demis Hassabis called for greater regulation to quell existential fears over tech with above-human levels of intelligence, stating, “We must take the risks of AI as seriously as other major global challenges, like climate change.
“It took the international community too long to coordinate an effective global response to this, and we’re living with the consequences of that now. We can’t afford the same delay with AI.”
Gary Gensler, the U.S. Securities and Exchange Commission Chair has reportedly stated, “that without some form of intervention, a financial crisis stemming from the widespread use of artificial intelligence is nearly unavoidable.”
With U.K. Prime Minister Rishi Sunak’s global AI Safety Summit on the 1st and 2nd of November at the iconic Bletchley Park – the home of Britain’s World War II Codebreakers, the Generative AI investment hype is playing out with both supporters and detractors alike, the makings for a blockbuster global summit.
At his summit speech, Sunak is expected to refer to the new opportunities for economic growth offered by powerful AI systems but will also acknowledge they bring new dangers including risks of cybercrime, designing of bioweapons, disinformation and upheaval to jobs.
Supporters will point to the star AI investment story, Nvidia, whose share price is up over 200 percent this year due to demand for its graphics processing unit (GPU) used in data centres to train and run AI models.
The hunt for the next Nvidia dominates investor forums and discussions. Mentions of AI product roadmaps have increased significantly on earnings announcements from publicly quoted companies, and warnings about not falling for the AI hype have surged in response.
Analysts are urging investors to focus on companies that generate revenue from AI or use it to gain a strategic edge and improve products, over companies that are pre-revenue or cash negative and talking up future values.
Is Generative AI All Hype?
For every winner there is a loser, and there will be plenty of losers in the generative AI investment craze.
Gary Marcus, Emeritus Professor of Psychology and Neural Science at NYU and a founder of the Center for the Advancement of Trustworthy AI, writes “The revenue isn’t there yet, and might never come. The valuations anticipate trillion-dollar markets, but the actual current revenues from Generative AI are rumoured to be in the hundreds of millions. Those revenues genuinely could grow by 1000x, but that’s mighty speculative.”
People who have used ChatGPT will have noticed that it can make mistakes and makes up facts simply because it does not have the right data. Tech companies are working to improve their models. Some generative AI models, such as Bing Chat or GPT-4 are providing footnotes with sources that tell users where the response is coming from and to check its accuracy.
For some potential customers, that accuracy will never be good enough. Craig Martell, chief digital and AI officer at the US. Defense Department recently told a conference, “I need five nines – 99.999 percent accuracy – of correctness.” Soldiers using AI cannot afford mistakes literally blowing up in their faces.
Financial services institutions are likely to be sceptical about the performance of Generative AI and its use cases in highly regulated businesses that require accuracy. Hedge fund quants using algorithmic and high frequency trading strategies will also set a high bar for Generative AI.
If the defence and highly regulated industries are unlikely to deploy Generative AI in the short to medium term for mission critical use cases due to its performance meeting accuracy targets, is the AI existential threat real or perceived?
Gary Ackerman, an associate professor and associate dean at the College of Emergency Preparedness, Homeland Security and Cybersecurity (CEHC) writes that one of the ways terrorists can get leverage is through an enabling technology like AI, by hacking into an existing AI that, for example, controlled nuclear weapons systems and set off nuclear war.
Political and business leaders must ensure claims of AIs existential threats to humanity are real and credible, and propose real and credible policies to help mitigate these threats or tamp down the narrative and put it into a more understandable context for citizens.
What is clear to many experts is that AI is positioning to unleash a new era of menacing threats from cybercriminals. Cybersecurity leaders point to additional emerging threats posed by AI, and more specifically Generative AI in volume and increased threats to cyber attack vectors in the future.
What is also clear is that AI is positioning to be the next big tech investment opportunity for savvy tech investors.
There Is Economic Substance
A study from McKinsey estimates that the impact of Generative AI on productivity could add trillions of dollars of value to the global economy. Its analysis put the potential added value at between $2.6 billion and $4.4 billion annually. That’s roughly equivalent to the entire GDP of leading European countries like the U.K, France, and Germany.
The report also suggests generative AI could enable labor productivity growth of 0.1 percent to 0.6 percent annually up to 2040 depending on how quickly the technology is adopted and how workers’ time is redeployed.
In the U.K. the Confederation of British Industry (CBI) estimates there are currently 3,011 companies classified as AI companies supporting a total of $11 billion in gross value added to the economy. That is roughly similar to the impact of the U.K. travel agency industry. The CBI report also highlights AI’s “potential to fundamentally change all aspects of our digital lives and raise the game on productivity for all sorts of businesses.”
Ash Patel of investment bank Deutsche Numis says: “The Generative AI sector shows AI’s rapid and explosive potential. Companies here are showing strong Product Market Fit (PMF) and ROI along with hyper scalability, which is driven by a truly exceptional crop of founders. We believe that Generative AI adoption will be quicker than, and catalysed by, cloud.”
It’s Not Just About Chasing Unicorns
Major public and private companies are investing in research and development (R&D) on AI solutions in major ways and the spending by tech market leaders is eye watering.
The stock market superstar Nvidia has increased R&D spending by 39 percent as they increase development of AI-focused chips. BT Group in the U.K. says it expects to replace up to 10,000 jobs with AI by 2030.
Meta plans to spend $33 billion this year to support “ongoing build-out of AI capacity.” Amazon is committed to spending many billions of dollars to build significant large language models.
EY has recently announced a $1 billion investment in a next-gen technology platform to enhance their Assurance Service Capabilities which includes AI software. Accenture announced it will invest $3 billion in AI to “accelerate clients’ reinvention” and will double AI talent to 80,000 people through hiring, acquisitions, and training.
Making The Investment Case For Generative AI
The creation of new jobs by AI with the replacement of old jobs with AI is one of the drivers for investment in the sector. Many believe Generative AI will substantially enhance productivity.
Even when it is imperfect AI can enhance efficiency and cut costs in many industries and particularly in sectors where the US Defense Department standard of “five nines” accuracy, also know as Six Sigma, is not essential. For many businesses, a “nine five” (95 percent, or Two Sigma) would knock the ball out the park on accuracy – of correctness, and radically increase quality and performance.
Many startups operate at One Sigma, 68 percent or below accuracy – of correctness, especially those that move fast and break things. This quality threshold is not suitable for highly regulated activities, like those in financial services. Keeping an eye on the AI startups moving up the funding ladder is key here. Maturing startups deliver greater quality products and services in C+ later stage rounds and are funded by more experienced investors.
The McKinsey study points out that Generative AI can be deployed today with a high impact on solving business problems faced by companies in Sales and Marketing and Customer Operations. There are business across a wide range of industries where generative AI’s data analysis in these area could be very useful, and correct and accurate enough for the job at hand.
There is also the future potential to consider. Generative AI will in time play a part in new services and businesses which have not been developed yet. AI enthusiasts make the comparison with the invention of electricity and generative AI as the new electricity sparking new ideas and services.
Investment bank Deutsche Numis propriety research and analysis of industry data by shows that in the second quarter of this year generative AI companies raised around $2.85 billion in funding. That is an increase of over $1.7 billion on the first quarter. The amount of money raised in the first half of 2023 is nearly five times higher than the whole of 2022.
The average funding deal size has grown too. So far this year it is around $61 million. That is more than double the average of $24 million in 2022 and more than treble the $18 million in 2021. So far, much of the AI investment has been in foundation models, which will likely serve as a catalyst for other areas including apps, machine learning operations, infrastructure and vertically integrated models.
Globally, AI companies (Generative and beyond) have raised over $354.5 billion in funding, with around $134 billion of this raised in the US followed by $67 billion from firms based in Asia, and $53 billion from those in Europe.
Deutsche Numis is working at the sharp end of fundraising in the sector and is forecasting further huge increases in the value of investments made in generative AI companies. It is confident that AI will create a vast number of highly successful companies on a scale that could mirror and potentially exceed the number of companies created from the rapid adoption of Cloud technologies.
Head For The Clouds
The distribution of cloud company market capitalization has highlighted AI’s significant potential for value expansion. Deutsche Numis forecasts that AI’s evolution will create a new cohort of highly disruptive, multi-billion-dollar companies, fueled by a further surge of investment into the sector.
A key point identified by its analysis is that later stage generative AI fund raising has only just begun. Around 42 percent of fund raising by the sector this year has been at seed and angel stage, with 25 percent at Series A and B, and just 8 percent has been Series C+.
Adjusting growth to listed peers illustrates how AI valuations may not be unreasonable. Although companies are nascent, the massive market opportunity for AI products and services, combined with rapid adoption and PMF position some AI companies to deliver value at unprecedented rates.
Picking winners and losers in the Generative AI space is likely to prove challenging. The sceptics might not be entirely wrong, but taking everything into account, investing in the Generative AI space appears to offer very attractive opportunities, if you can follow and keep up with the smart money.